As medical technology continues to advance, patients face an increasing variety of treatment choices. Further, the amount of time that physicians have to listen to patients and tailor treat-ments to patient preferences is decreasing due to changes in the way medicine is being practiced. Therefore, it is especially important to find new ways to increase both the objectivity and the degree of patient participation in treatment decisions. In this proposal we describe a three-part process to achieve these goals and disseminate the results over the World Wide Web (WWW). In the first part of our study, we will characterize patient preferences for common multi-dimensional health states in rheumatoid arthritis (RA) using Time Trade-Off and Visual Analog Scale methods. We will study patient values in 300 RA patients recruited over the WWW. We will learn about how patients value losses and gains in different dimensions of health, and whether disability and pain caused by drug toxicities deserve greater weight in decision-making than similar outcomes caused by disease progression. We will also compare patients' values for outcomes to those of a sample of rheumatologists. In the second part of our study, we will develop a decision model for evaluating different treatments for rheumatoid arthritis and integrate this model into a second computer program. This program will be able to describe health outcomes, measure preferences for those outcomes, and then generate individual specific treatment recommendations. We will validate this program by comparison of its therapeutic recommendations to those of expert rheumatologists. In the third part of this study, we will apply the program in a representative sample of rheumatoid arthritis patients to study the degree to which patients' current therapies for RA are consistent with their preferences, to assess patient reactions to the program, and to further validate the program's recommendations. When we have perfected our model and are finished with validation studies, we will make our software available over the WWW to RA patients and rheumatologists who wish to use this software for decision support. We will also distribute generalized versions of the software for use to other investigators in other disease areas via the WWW site.